You have production datacenters, you have a number of servers that are direct or network attached as primary storage.  You need to have disaster recovery plan. You would do this through Backup and replication.  physical and virtual servers or Apps. On-perm or cloud ? or both ?  Restore or compete automated Failover and failback ?

These are some questions you need to answer . ..

It’s an automatic failover. You can backup to your primary storage or backup to managed backup appliance. The provider backs up the physical and virtual machine images and its data to the cloud, and it boots up from restored VMs. The managed backup appliances is talking to the cloud provider, and that’s another way to look at that from a production datacenter point of view.

Multiple Replication options exist for DRaaS

Multiple replications often exists for DR as a service and this is something you have to take into consideration. SAN-to-SAN replication actually require matching hardware. Host replication, and they would be agent based as well as the hypervisor replication and those are actually replicating constantly to other hosts.

When you’re talking about actual replication from a software perspective, software versus hardware replication; in software replication you don’t have to actually go from one datacenter where your hardware is one standard. You may be all dell but you are going over to another datacenter and that datacenter doesn’t have to have dell. The software is obscuring that layer from the DR software. You don’t have to create a second datacenter that happens to be in a cloud managed by somebody else.

There’s lots of different flavors to the stuff, different ways to look at it quite frankly depending on how often you need to update, how often you want asynchronous or synchronous, how much out of sync you are allowed to be. It kind of determines how much money you want to spend will determine which of these you are actually going to go after.